Before you buy an ERP: eight points that decide whether it works

When an ERP project fails, the usual explanation is "the system wasn't right for us". In practice, most of the failures we see trace back to decisions settled — or not settled — before the contract was signed.

1. Is your document cycle written down? Before looking at any proposal, write what actually happens: who raises the purchase order, who approves it, when stock enters the warehouse, when the journal entry is posted. If you cannot write that cycle down, the system will impose its own.

2. Who owns the decision? An ERP project without one decision owner becomes an endless negotiation between departments. Name one person with authority to settle disputes.

3. Which reports will decisions be made on? List the reports management actually needs. That list shapes the data model far more than a feature comparison does.

4. What state is your current data in? Duplicate items, customers under several names, balances that don't reconcile — all of it travels with you unless it is cleaned. Cleaning before migration is far cheaper than cleaning after.

5. Have you settled the chart of accounts and cost centres? Changing the chart of accounts after go-live is painful. Sit with your accountant and settle it early.

6. Who is trained, and when? Training crammed into the final week before launch guarantees user resistance. Make it part of the schedule, not an appendix.

7. Will you run in parallel? Running alongside the old system for a limited period exposes discrepancies before they become an accounting problem. It costs time, but far less than a wrong close.

8. What happens after go-live? Launch is not the end of the project. Decide who supports users through the first closing cycle, who holds change authority, and how changes are requested.

In short The system is a tool. What determines the outcome is the clarity of the process before the tool. Answer these eight points honestly and you are ready to compare proposals on a real basis.